Pile foundation block construction system suitable for karst cave area

Abstract

The invention belongs to the technical field of geotechnical engineering and geological engineering, and particularly relates to a pile foundation block construction system suitable for a karst cave area, which comprises a data acquisition and processing system. According to the pile foundation block construction system suitable for the karst cave area, the data collecting and processing system is arranged, and the seismic source module, the unmanned aerial vehicle technology and the high-precision sensor are integrated, so that the recognition precision of a karst cave is remarkably improved, the geological heterogeneity is accurately detected, and reliable data support is provided for pile foundation design. Meanwhile, the pile foundation design method based on intelligent analysis avoids the limitation of traditional empirical design, optimizes the pile foundation layout scheme, and improves the bearing capacity of the pile foundation. In the construction process, the intelligent monitoring and hydraulic deviation correcting mechanism can adjust the deviation of the pile foundation in real time, the safety and efficiency of the construction process are improved, and therefore the problems that an existing pile foundation project in the karst cave area is high in accident rate and low in adaptability and efficiency are solved.

Description

Technical Field

[0001] This invention relates to the fields of geotechnical engineering and geological engineering, and in particular to a block-based construction system for pile foundations suitable for karst cave areas. Background Technology

[0002] In areas with severe karst cave development, the karst geology is characterized by complex morphology, random spatial distribution, and variable filling materials, leading to multiple technical challenges in pile foundation construction: I. Traditional geophysical exploration methods (such as high-density electrical resistivity tomography and single-frequency surface wave exploration) are limited by resolution and detection depth, resulting in a low identification rate and a very high rate of missed detection for multi-layered nested caves and hidden cavities.

[0003] Second, conventional pile foundation design uses uniform pile layout plus empirical safety factor, without considering the heterogeneity of karst cave space, resulting in redundant pile length or insufficient local bearing capacity, and increased project cost.

[0004] Third, accidents such as collapse of the cave roof, deviation of pile holes, and sudden surge of groundwater occur frequently.

[0005] Traditional pile foundation design and construction methods often fail to meet safety and economic requirements. Therefore, considering the characteristics of karst caves, a block-based construction system for pile foundations suitable for karst cave areas is proposed to improve construction adaptability and efficiency while ensuring construction safety. Summary of the Invention

[0006] Based on the existing technical problems of high accident rate and low adaptability and efficiency of pile foundation engineering in karst cave areas, this invention proposes a block-based construction system for pile foundations suitable for karst cave areas.

[0007] The present invention proposes a block-based construction system for pile foundations suitable for karst cave areas, including a data acquisition and processing system, which consists of a ground data acquisition module, an unmanned aerial vehicle (UAV) data acquisition module, and a cloud processing platform.

[0008] Both the ground data acquisition module and the UAV data acquisition module are connected to the cloud processing platform via 5G data.

[0009] The ground data acquisition module consists of a seismic source module, a sensing module connected to the seismic source module, and a control module connected to the sensing module.

[0010] Preferably, the seismic source module includes a hydraulic vibrator, a low-frequency air gun, and a micro-motion array. The hydraulic vibrator is used to generate a 5-15Hz frequency band detection surface wave, the low-frequency air gun is combined with ambient noise to generate a 1-5Hz frequency band detection surface wave, and the micro-motion array is used to generate a 0.1-1Hz frequency band detection surface wave.

[0011] Preferably, the sensing module is used to monitor the high-frequency data generated by the seismic source module and feed it back to the control module.

[0012] The sensing module includes a high-frequency detector for monitoring the hydraulic vibrator, a broadband seismograph for monitoring the ambient noise and the low-frequency air gun, and an ultra-low frequency seismograph for monitoring the micro-motion array.

[0013] Preferably, the sensing module transmits the sensed and detected data to the cloud processing platform through the control module. The control module is used to process and communicate the sensed and detected data by the sensing module. The control module includes an edge computing gateway, a 5G communication module, and a hydraulic correction mechanism.

[0014] The hydraulic correction mechanism is used to dynamically adjust the verticality of the pile foundation in response to control commands.

[0015] The hydraulic correction mechanism consists of a high-pressure hydraulic pump station, a double-acting hydraulic cylinder, a servo proportional valve, an tilt sensor, and a displacement sensor.

[0016] Preferably, the cloud processing platform is used to process the data monitored and sensed by the sensing module. The cloud processing platform processes the monitoring data of the ground data acquisition module and the UAV data acquisition module through a finite element-machine learning hybrid model, and provides optimized parameter recommendations for the control module based on the predicted control effect.

[0017] The UAV data acquisition module consists of an image acquisition module, a lidar, an environmental monitoring module, and a 5G communication module installed on the UAV.

[0018] The image acquisition module uses a high-definition camera or an infrared camera, and the environmental monitoring module consists of a temperature and humidity sensor and a wind speed sensor.

[0019] Preferably, a block-based construction system for pile foundations suitable for karst cave areas includes the following steps: Step 1: Multi-band surface wave collaborative detection and data acquisition. Three-band surface wave collaborative detection of 5-15Hz, 1-5Hz, and 0.1-1Hz, combined with UAV surveying, was used to obtain multi-scale karst cave data.

[0020] Step 2: Spatiotemporal synchronization and data preprocessing. Based on a GPS / BeiDou dual-mode clock synchronizer, time synchronization of multiple devices is achieved with a synchronization error of ≤1μs, ensuring the consistency of time sequence between ground and UAV data.

[0021] Step 3: Random Forest Classification and Block Division. A random forest classifier is used to perform four-dimensional feature analysis on the thickness of the cave roof, shear wave velocity, horizontal distance, and collapse probability to divide the cave into Class I / II design blocks.

[0022] Input features include the thickness of the cave roof slab δ, shear wave velocity Vs, horizontal distance L, collapse probability Pc, cave morphology factor K = volume / surface area, and wave velocity gradient ΔVs / Δh.

[0023] Data fusion decision: Pc = α·P 地面 +β·P 无人机 +γ·P 气象 Where α, β, and γ are dynamic weighting coefficients that are adjusted in real time according to the data confidence level.

[0024] Step 4: Edge-cloud collaborative control and dynamic regulation. Implement multi-parameter feedback control based on the edge-cloud collaborative architecture, in which a lightweight anomaly detection model is deployed at the edge layer with a response latency of <100ms.

[0025] Preferably, in step one, a drone is first dispatched to conduct a preliminary on-site survey to obtain preliminary topographic data of the karst cave area. Environmental monitoring points are set up around the construction area to continuously acquire meteorological and environmental data, providing basic data for risk management during the construction process. Then, a signal is excited by a hydraulic servo vibrator, and a high-frequency detector array is used to collect the data. The signal-to-noise ratio is improved by multiple superpositions to obtain surface wave data in the 5-15Hz frequency band.

[0026] In step one, broadband seismometers are deployed in an L-shaped array with an aperture of 50-200m. Surface wave dispersion curves are extracted using environmental noise cross-correlation technology to obtain surface wave data in the 1-5Hz frequency band.

[0027] In step one, the micro-motion array is continuously observed for ≥24 hours, and 3-5 sub-arrays of ultra-low frequency seismometer are set up with a spacing of 100-500m to acquire surface wave data in the 0.1-1Hz frequency band.

[0028] In step one, a drone flies above the seismic source module, generates a DTM through 3D scanning with lidar, captures the cave topography with a high-definition camera, identifies thermal anomalies with an infrared camera, and monitors meteorological data in real time through temperature and humidity sensors and wind speed sensors to assist in risk assessment.

[0029] Preferably, the block division rules in step three are as follows: In Class I areas, the depth of the pile tip embedded in intact bedrock is ≥ 3 times the pile diameter.

[0030] Pre-grouting volume in Zone II: Q = πD²h·η Where π is the mathematical constant pi, used to calculate the cross-sectional area of ​​the pile foundation. η is the karst cave filling rate, h is the karst cave height, and D is the design diameter of the pile foundation.

[0031] Pre-grouting volume after risk dynamic correction .

[0032] Where k is the risk correction coefficient, a parameter that adjusts the impact of collapse probability on grouting volume; a larger value indicates higher sensitivity to collapse probability. Pc is the collapse probability, the probability of the karst cave collapsing during construction, calculated using a random forest model; a larger value indicates higher risk.

[0033] Preferably, step four includes hardware deployment, cloud platform intelligent decision-making, and a digital twin system. The hardware deployment uses an NVIDIA Jetson NX device to run a lightweight MobileNetV3 model with a model size of <500KB, which can be run on the NVIDIA Jetson NX device to detect pile foundation displacement and karst cave collapse risks in real time.

[0034] The cloud platform's intelligent decision-making includes LSTM-GRU hybrid neural network prediction of grouting volume requirements and encrypted data sharing across multiple construction sites, with a model update cycle of ≤24h.

[0035] Preferably, the digital twin system includes a BIM model fused with ground-penetrating radar data, a real-time refresh rate of 60Hz, and employs a grouting pressure-flow dual-loop PID controller with an integral time constant Ti=0.8s and a pile foundation verticality fuzzy control system.

[0036] Step four also includes a conflict arbitration mechanism, which uses confidence-weighted averages when there are discrepancies between ground and drone data.

[0037] The beneficial effects of this invention are as follows: 1. By setting up a data acquisition and processing system and integrating seismic source modules, UAV technology, and high-precision sensors, the accuracy of karst cave identification was significantly improved, and geological heterogeneity was accurately detected, providing reliable data support for pile foundation design. Simultaneously, the intelligent analysis-based pile foundation design method avoids the limitations of traditional experience-based design, optimizes the pile foundation layout scheme, and enhances the bearing capacity of the pile foundation. During construction, intelligent monitoring and hydraulic correction mechanisms can adjust pile foundation deviations in real time, improving the safety and efficiency of the construction process, thereby solving the problems of high accident rates and low adaptability and efficiency in existing pile foundation projects in karst cave areas.

[0038] 2. By setting steps one through four, when using the system, the LiDAR surface model and surface wave inversion profile are embedded into the same BIM scene to generate a 3D model of the karst cave. Then, the thermal anomaly boundary of the UAV infrared image is extracted and superimposed with the ground wave velocity anomaly area to identify seepage channels. Finally, the dynamic grouting parameters are output by integrating the weights of multi-source data through a random forest model. This achieves the effect of full-process monitoring and control of the collection, fusion, analysis and regulation of block-based construction data for pile foundations in karst cave areas, thereby solving the problems of high accident rate and low adaptability and efficiency of existing pile foundation projects in karst cave areas. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of a block-based construction system for pile foundations suitable for karst cave areas proposed in this invention; Figure 2 This is a schematic diagram of the ground data acquisition module structure of a pile foundation block construction system suitable for karst cave areas proposed in this invention; Figure 3 This is a schematic diagram of the UAV data acquisition module structure of a pile foundation block construction system suitable for karst cave areas proposed in this invention; Figure 4 This invention presents a cloud-based data processing flowchart of a block-based pile foundation construction system suitable for karst cave areas. Figure 5 This invention presents a flowchart illustrating the structure and data processing of a ground data acquisition module for a block-based pile foundation construction system suitable for karst cave areas. Figure 6 This is a schematic diagram of a two-dimensional shear wave velocity profile A for a block-based construction system for pile foundations in karst areas proposed in this invention; Figure 7 This is a schematic diagram of a two-dimensional shear wave velocity profile B for a block-based construction system for pile foundations in karst areas proposed in this invention; Figure 8 This is a schematic diagram of a three-dimensional shear wave velocity model for a block-based construction system for pile foundations in karst cave areas proposed in this invention. Figure 9 This is a schematic diagram of advanced detection data processing and model deduction iterative optimization of a block-based pile foundation construction system suitable for karst cave areas proposed in this invention; Figure 10 This is a schematic diagram (B) illustrating the advanced detection data processing and model deduction iterative optimization of a block-based pile foundation construction system suitable for karst cave areas proposed in this invention.

[0040] In the diagram: 1. Data acquisition and processing system; 101. Ground data acquisition module; 1011. Seismic source module; 101101. Hydraulic vibrator; 101102. Low-frequency air gun; 101103. Micro-motion array; 1012. Sensing module; 101201. High-frequency detector; 101202. Broadband seismometer; 101203. Ultra-low frequency seismometer; 1013. Control module; 101301. Edge computing gateway; 101302. 5G communication module one; 101303. Hydraulic correction mechanism; 102. UAV data acquisition module; 1021. Image acquisition module; 1022. LiDAR; 1023. Environmental monitoring module; 1024. 5G communication module two; 103. Cloud processing platform. Detailed Implementation

[0041] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0042] Reference Figures 1-3 A modular construction system for pile foundations suitable for karst cave areas includes a data acquisition and processing system 1, which consists of a ground data acquisition module 101, a UAV data acquisition module 102, and a cloud processing platform 103.

[0043] Both the ground data acquisition module 101 and the UAV data acquisition module 102 are connected to the cloud processing platform 103 via 5G data.

[0044] The ground data acquisition module 101 consists of a seismic source module 1011, a sensing module 1012 connected to the seismic source module 1011, and a control module 1013 connected to the sensing module 1012.

[0045] The seismic source module 1011 includes a hydraulic exciter 101101, a low-frequency air gun 101102, and a micro-motion array 101103. The hydraulic exciter 101101 is used to generate a detection surface wave in the 5-15Hz frequency band. The low-frequency air gun 101102, combined with ambient noise, is used to generate a detection surface wave in the 1-5Hz frequency band. The micro-motion array 101103 is used to generate a detection surface wave in the 0.1-1Hz frequency band.

[0046] Furthermore, the sensing module 1012 is used to monitor the high-frequency data generated by the source module 1011 and feed it back to the control module 1013.

[0047] The sensing module 1012 includes a high-frequency detector 101201 for monitoring the hydraulic vibrator 101101, a broadband seismograph 101202 for monitoring environmental noise and the low-frequency air gun 101102, and an ultra-low frequency seismograph 101203 for monitoring the micro-motion array 101103.

[0048] The sensing module 1012 transmits the sensing and detection data to the cloud processing platform 103 through the control module 1013. The control module 1013 is used to process and communicate the sensing data monitored and perceived by the sensing module 1012. The control module 1013 includes an edge computing gateway 101301, a 5G communication module 101302, and a hydraulic correction mechanism 101303.

[0049] Furthermore, the hydraulic correction mechanism 101303 is used to dynamically adjust the verticality of the pile foundation in response to control commands.

[0050] Specifically, the hydraulic correction mechanism 101303 consists of a high-pressure hydraulic pump station, a double-acting hydraulic cylinder, a servo proportional valve, an inclination sensor, and a displacement sensor.

[0051] The high-pressure hydraulic pump station adopts the Bosch Rexroth SYDFE series electro-hydraulic servo system, with an output pressure ≥21MPa and an adjustable flow rate range of 5-50L / min. It is equipped with an accumulator to ensure instantaneous high flow rate requirements. The double-acting hydraulic cylinders use the Parker HMI series, with a thrust ≥50kN (single cylinder), a stroke of 100-300mm, and a repeatability of ±0.1mm. Four to eight sets of hydraulic cylinders are symmetrically arranged to form a multi-degree-of-freedom correction system (horizontal X / Y axis + rotational compensation). The servo proportional valves use the Moog D633 series, with a response time <10ms, supporting PID closed-loop control and direct communication with the 101301 edge computing gateway (such as NVIDIA Jetson NX). The tilt sensor uses the SICK DBS36E, with a range of ±10° and an accuracy of ±0.01°, to monitor the verticality of the pile in real time. The displacement sensor uses the MTS Temposonics magnetostrictive linear sensor, with a range of ±150mm and a resolution of 1μm, to provide feedback on the actual displacement of the hydraulic cylinder.

[0052] The working principle of the hydraulic correction mechanism 101303 is as follows: the tilt sensor collects the tilt angle of the pile in real time, the displacement sensor monitors the stroke of the hydraulic cylinder, the edge computing gateway 101301 runs the MobileNetV3 model to analyze the data and generate correction commands, the servo proportional valve receives the PWM control signal and adjusts the oil flow and direction of the hydraulic cylinder, the double-acting hydraulic cylinder pushes the pile foundation guide frame to apply the reverse correction force, dynamically adjust the verticality of the pile, and adopts the fuzzy PID control algorithm combined with the pile-soil interaction model to optimize the thrust curve and avoid over-adjustment.

[0053] The cloud processing platform 103 is used to process the data monitored and sensed by the sensing module 1012. The cloud processing platform 103 processes the monitoring data of the ground data acquisition module 101 and the UAV data acquisition module 102 through a finite element-machine learning hybrid model, and provides optimized parameter recommendations to the control module 1013 based on the predicted control effect.

[0054] The UAV data acquisition module 102 consists of an image acquisition module 1021, a lidar 1022, an environmental monitoring module 1023, and a 5G communication module 1024 installed on the UAV.

[0055] The image acquisition module 1021 uses a high-definition camera or an infrared camera, and the environmental monitoring module 1023 consists of a temperature and humidity sensor and a wind speed sensor.

[0056] Furthermore, during use, the high-definition camera or infrared camera mounted on the drone is used to take aerial photos of the construction area to obtain topographic images and video data of the karst cave area. The drone is equipped with a LiDAR 1022 to perform a three-dimensional scan of the karst cave area and generate an accurate digital terrain model. At the same time, the drone is equipped with temperature and humidity sensors and wind speed sensors to monitor the meteorological conditions at the construction site in real time.

[0057] Furthermore, during the detection process, the drone flies above the seismic source module 1011 to collect high-precision data. The collected images, LiDAR data, and seismic source fluctuation data are transmitted in real time to the cloud processing platform 103 via the 5G communication module 1024 to ensure immediate feedback of information.

[0058] By setting up a data acquisition and processing system 1, and integrating a seismic source module 1011, UAV technology, and high-precision sensors, the accuracy of karst cave identification was significantly improved, and geological heterogeneity was accurately detected, providing reliable data support for pile foundation design. Simultaneously, the intelligent analysis-based pile foundation design method avoids the limitations of traditional experience-based design, optimizes the pile foundation layout scheme, and enhances the bearing capacity of the pile foundation. During construction, the intelligent monitoring and hydraulic correction mechanism 101303 can adjust pile foundation deviations in real time, improving the safety and efficiency of the construction process, thereby solving the problems of high accident rates and low adaptability and efficiency in existing pile foundation projects in karst cave areas.

[0059] Reference Figures 1-10 A modular construction system for pile foundations suitable for karst cave areas includes the following steps: Step 1: Multi-band surface wave collaborative detection and data acquisition. Using three-band surface wave collaborative detection (5-15Hz, 1-5Hz, 0.1-1Hz) and UAV-assisted surveying, multi-scale karst cave data is obtained. Specifically, in step one, drones are first dispatched to conduct preliminary on-site surveys to obtain preliminary topographic data of the karst cave area. Environmental monitoring points are set up around the construction area to continuously acquire meteorological and environmental data, providing basic data for risk management during the construction process. Then, signals are excited by hydraulic servo vibrators and collected by a 101201 array of high-frequency detectors. The signal-to-noise ratio is improved by multiple superpositions to obtain surface wave data in the 5-15Hz frequency band.

[0060] Specifically, the hydraulic servo vibrator has an adjustable output frequency of 5-50Hz and a peak output force of ≥5 tons. It is used to excite high-frequency surface wave signals and improve the signal-to-noise ratio through multiple excitations (such as 10 superpositions).

[0061] The 101201 high-frequency geophone employs a high-frequency seismic geophone with a sensitivity ≥0.8V / m / s, suitable for shallow, high-resolution data acquisition. Its dense arrangement at 2-5m satisfies the spatial sampling theorem.

[0062] In step one, broadband seismometers 101202 are deployed in an L-shaped array with an aperture of 50-200m. Surface wave dispersion curves are extracted using environmental noise cross-correlation technology to obtain surface wave data in the 1-5Hz frequency band.

[0063] Specifically, the seismic source equipment is a passive environmental noise source combined with an active supplementary low-frequency air gun 101102 or a heavy hammer vibration device. Natural or construction environmental noise (such as mechanical vibration, vehicle traffic) is used as the seismic source. Continuous vibration signals are collected by a broadband seismograph array. The broadband seismograph is a GURALP CMG-3T with a frequency band of 0.1-100Hz and a dynamic range of ≥140dB. It adopts a circular or L-shaped array with an aperture of 50-200m. The surface wave dispersion curve is extracted by noise cross-correlation technology.

[0064] In step one, the micro-motion array 101103 is continuously observed for ≥24 hours, and 3-5 sub-arrays of the ultra-low frequency seismograph 101203 are set up with a spacing of 100-500m to acquire surface wave data in the 0.1-1Hz frequency band.

[0065] Specifically, the seismic source equipment uses the 101103 micro-motion array technology in conjunction with a heavy-duty controllable seismic source, utilizing natural micro-vibrations (such as atmospheric disturbances) as an ultra-low frequency seismic source. The heavy-duty controllable seismic source uses the Vibroseis controllable seismic source vehicle, with a minimum frequency of 0.1Hz.

[0066] The 101203 ultra-low frequency seismograph uses the STS-2, with a frequency band of 0.008-50Hz. It features a low-noise substrate (<1nm / s²) and is deployed in 3-5 sub-arrays, each containing 6-12 sensors with a spacing of 100-500m.

[0067] In step one, a drone flies above the seismic source module 1011, and a DTM is generated by three-dimensional scanning using a lidar 1022. A high-definition camera captures the cave topography, an infrared camera identifies thermal anomalies, and meteorological data is monitored in real time by temperature and humidity sensors and wind speed sensors to assist in risk assessment.

[0068] Step 2: Spatiotemporal synchronization and data preprocessing. Based on a GPS / BeiDou dual-mode clock synchronizer, time synchronization of multiple devices is achieved with a synchronization error of ≤1μs, ensuring the consistency of time sequence between ground and UAV data.

[0069] Specifically, a rubidium clock and a temperature-compensated crystal oscillator GPS / BeiDou dual-mode atomic clock are used to provide a reference clock. Device-level synchronization is achieved based on the PTPv2 protocol. The UAV LiDAR point cloud and the ground detection grid are aligned using the ICP algorithm, and timestamp interpolation is used to eliminate transmission delays, achieving spatiotemporal registration. Data at different resolutions (LiDAR centimeter-level and surface wave meter-level) are downsampled or interpolated to achieve scale matching and data normalization.

[0070] Step 3: Random forest classification and block division. The random forest classifier is used to perform four-dimensional feature analysis on the thickness of the cave roof, shear wave velocity, horizontal distance and collapse probability to divide the cave into Class I / II design blocks. Specifically, the input features are: cave roof thickness δ, shear wave velocity Vs, horizontal distance L, collapse probability Pc, cave morphology factor K = volume / surface area, and wave velocity gradient ΔVs / Δh.

[0071] Data fusion decision: Pc = α·P 地面 +β·P 无人机 +γ·P 气象 Wherein, α, β, γ are dynamic weighting coefficients that are adjusted in real time according to the data confidence level. In this embodiment, UAV meteorological data accounts for 15-20%, and ground wave velocity parameters account for 50-55%.

[0072] Furthermore, the block division rules in step three are as follows: In Class I areas, the depth of the pile tip embedded in intact bedrock is ≥ 3 times the pile diameter; Pre-grouting volume in Zone II: Q = πD²h·η Where π is the mathematical constant pi, used to calculate the cross-sectional area of ​​the pile foundation. η is the karst cave filling rate, h is the karst cave height, and D is the design diameter of the pile foundation.

[0073] Pre-grouting volume after risk dynamic correction ; Where k is the risk correction coefficient, a parameter that adjusts the impact of collapse probability on grouting volume; a larger value indicates higher sensitivity to collapse probability. Pc is the collapse probability, the probability of the karst cave collapsing during construction, calculated using a random forest model; a larger value indicates higher risk.

[0074] Furthermore, in the specific implementation process, the collapse probability Pc is updated in real time through a machine learning model, and the coefficient k is determined through historical data inversion or field experiments. When Pc approaches 1, a manual review mechanism is initiated to avoid model overfitting.

[0075] Step 4: Edge-cloud collaborative control and dynamic regulation. Implement multi-parameter feedback control based on the edge-cloud collaborative architecture, in which a lightweight anomaly detection model is deployed at the edge layer with a response latency of <100ms.

[0076] Furthermore, step four includes hardware deployment, cloud platform intelligent decision-making, and a digital twin system. Among these, the hardware deployment involves running a lightweight MobileNetV3 model on an NVIDIA Jetson NX device. The model size is less than 500KB and can be run on the NVIDIA Jetson NX device to detect the risk of pile foundation displacement and karst cave collapse in real time.

[0077] The cloud platform's intelligent decision-making includes LSTM-GRU hybrid neural network prediction of grouting volume requirements and encrypted data sharing across multiple construction sites, with a model update cycle of ≤24 hours.

[0078] The digital twin system includes a BIM model fused with ground-penetrating radar data, a real-time refresh rate of 60Hz, and employs a grouting pressure-flow dual-loop PID controller with an integral time constant Ti=0.8s and a pile foundation verticality fuzzy control system.

[0079] Step four also sets up a conflict arbitration mechanism. When there is a conflict between ground and UAV data, a confidence weighting is used. In this embodiment, LiDAR data has a weight of 40% and surface wave data has a weight of 60% for decision-making.

[0080] By setting steps one through four, during use, the LiDAR surface model and surface wave inversion profile are embedded into the same BIM scene to generate a 3D model of the karst cave. Then, the thermal anomaly boundary of the UAV infrared image is extracted and superimposed with the ground wave velocity anomaly area to identify seepage channels. Finally, the dynamic grouting parameters are output by integrating the weights of multi-source data through a random forest model. This achieves the effect of full-process monitoring and control of the collection, fusion, analysis and regulation of block-based construction data for pile foundations in karst cave areas, thereby solving the problems of high accident rate and low adaptability and efficiency of existing pile foundation projects in karst cave areas.

[0081] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A pile foundation block construction system suitable for karst area, comprising a data acquisition and processing system (1), characterized in that: The data acquisition and processing system (1) is composed of a ground data acquisition module (101), a UAV data acquisition module (102) and a cloud processing platform (103); The ground data acquisition module (101) and the UAV data acquisition module (102) are connected with the cloud processing platform (103) through 5G data; The ground data acquisition module (101) is composed of a seismic source module (1011), a perception module (1012) connected with the seismic source module (1011) and a control module (1013) connected with the perception module (1012).

2. A pile foundation blockization construction system suitable for use in a karst region according to claim 1, characterized in that: The seismic source module (1011) includes a hydraulic vibrator (101101), a low-frequency air gun (101102) and a micro-motion platform array (101103), the hydraulic vibrator (101101) is used to generate 5-15Hz frequency band surface wave, the low-frequency air gun (101102) is combined with environmental noise to generate 1-5Hz frequency band surface wave, and the micro-motion platform array (101103) is used to generate 0.1-1Hz frequency band surface wave.

3. A pile foundation blockization construction system suitable for use in a karst region according to claim 2, characterized in that: The perception module (1012) is used to monitor the high-frequency data generated by the seismic source module (1011) and feedback to the control module (1013); The perception module (1012) includes a high-frequency detector (101201) for monitoring the hydraulic vibrator (101101), a wide-frequency seismometer (101202) for monitoring the environmental noise and the low-frequency air gun (101102), and an ultra-low-frequency seismometer (101203) for monitoring the micro-motion platform array (101103).

4. A pile foundation blockization construction system suitable for use in a karst region according to claim 3, characterized in that: The perception module (1012) transmits the data detected by the perception module (1012) to the cloud processing platform (103) through the control module (1013), the control module (1013) is used to process and communicate the data monitored by the perception module (1012), and the control module (1013) includes an edge computing gateway (101301), a 5G communication module (101302) and a hydraulic deviation correction mechanism (101303); The hydraulic deviation correction mechanism (101303) is used to dynamically adjust the verticality of the pile foundation and respond to the control instruction; The hydraulic deviation correction mechanism (101303) is composed of a high-pressure hydraulic pump station, a double-acting hydraulic cylinder, a servo proportional valve, an inclination sensor and a displacement sensor.

5. A pile foundation blockization construction system suitable for use in a karst region according to claim 4, characterized in that: The cloud processing platform (103) is used to process the data monitored by the perception module (1012), the cloud processing platform (103) processes the monitoring data of the ground data acquisition module (101) and the UAV data acquisition module (102) through a finite element-machine learning hybrid model, predicts the control effect, and provides optimization parameter recommendation to the control module (1013); The UAV data acquisition module (102) is composed of an image acquisition module (1021), a laser radar (1022), an environment monitoring module (1023) and a 5G communication module (1024) installed on a UAV; The image acquisition module (1021) adopts a high-definition camera or an infrared camera, and the environment monitoring module (1023) is composed of a temperature and humidity sensor and a wind speed sensor.

6. A pile foundation blockization construction system suitable for use in a karst region according to claim 5, wherein It comprises the following steps: Step one, multi-frequency surface wave cooperative detection and data acquisition, 5-15Hz, 1-5Hz, 0.1-1Hz three frequency band surface wave cooperative detection and unmanned aerial vehicle cooperation survey, obtain multi-scale cave data; Step two, time and space synchronization and data preprocessing, based on GPS / Beidou dual-mode clock synchronizer to realize multi-device time synchronization, synchronization error ≤1μs, to ensure the time sequence consistency of ground and unmanned aerial vehicle data; Step three, random forest classification and block division, through random forest classifier to analyze the four-dimensional features of karst cave roof thickness, shear wave velocity, horizontal distance and collapse probability, and divide the design blocks of class I / II; Input features, karst cave roof thickness δ, shear wave velocity Vs, horizontal distance L, collapse probability Pc, karst cave morphology factor K = volume / surface area, wave velocity gradient ΔVs / Δh; Data fusion decision: Pc = a · P 地面 + β · P 无人机 + γ · P 气象 Wherein, α, β, γ are dynamic weight coefficients, which are adjusted in real time according to data confidence; Step four, edge-cloud collaborative control and dynamic regulation, based on edge-cloud collaborative architecture to implement multi-parameter feedback control, wherein the edge layer deploys a lightweight anomaly detection model, and the response delay is less than 100ms.

7. A pile foundation blockization construction system suitable for use in a karst region according to claim 6, characterized in that: In step one, first send the unmanned aerial vehicle to conduct preliminary site survey and obtain preliminary topographic data of the cave area, set up environmental monitoring points around the construction area, continuously obtain meteorological and environmental data, and provide basic data for risk management in the construction process, then excite signals through a hydraulic servo exciter, an array of high-frequency detectors (101201) collects, and the signal-to-noise ratio is improved through multiple superposition to obtain 5-15Hz band surface wave data; In step one, the wideband seismometer (101202) is arranged in an L-shaped array with an aperture of 50-200m, and the surface wave dispersion curve is extracted by using the environmental noise cross-correlation technology to obtain 1-5Hz band surface wave data; In step one, the micro-motion station array (101103) continuously observes for ≥24 hours, and the ultra-low frequency seismometer (101203) is arranged in 3-5 sub-station arrays with a spacing of 100-500m to obtain 0.1-1Hz band surface wave data; In step one, the unmanned aerial vehicle flies above the seismic source module (1011), the DTM is generated by three-dimensional scanning through the laser radar (1022), the high-definition camera takes pictures of the cave terrain, the infrared camera identifies thermal anomalies, and the temperature and humidity sensor and the wind speed sensor monitor the meteorological data in real time to assist risk assessment.

8. A pile foundation blockization construction system suitable for use in a karst region according to claim 7, characterized in that: In step three, the block division rules are: Class I region: pile end embedded in complete bedrock depth ≥3 times pile diameter; Class II region: pre-grouting amount Q=πD²h·η Wherein, π is the circular constant, used to calculate the cross-sectional area of the pile foundation; η is the filling rate of the cave, h is the height of the cave, and D is the design diameter of the pile foundation; Risk dynamic revision of pre-grouting amount ; Wherein, k is the risk correction coefficient, which adjusts the parameter of the influence of collapse probability on grouting amount; Pc is the collapse probability, which is the probability of cave collapse during construction, calculated by the random forest model, and the larger the value, the higher the risk.

9. A pile foundation blockization construction system suitable for use in a karst region according to claim 8, characterized in that: The step four includes hardware deployment, cloud platform intelligent decision and digital twin system, wherein the hardware deployment is that an NVIDIA Jetson NX device runs a light MobileNetV3 model with a model size < 500 KB and supports running on the NVIDIA Jetson NX device to detect pile foundation deviation and karst collapse risk in real time; The cloud platform intelligent decision includes LSTM-GRU hybrid neural network prediction of grouting quantity demand and multi-site data encryption sharing with a model update cycle ≤ 24 h.

10. A pile foundation blockization construction system suitable for use in a karst region according to claim 9, characterized in that: The digital twin system includes BIM model fusion with geological radar data, real-time refresh rate 60 Hz, and grouting pressure-flow double loop PID controller with an integral time constant Ti = 0.8 s, and pile foundation verticality fuzzy control system; The step four further sets a conflict arbitration mechanism, and when the ground and unmanned aerial vehicle data are contradictory, confidence weighted is adopted.

Images